Significance and timing of the mid-17th-century eruption of Long Island,Papua New Guinea

Abstract

Tibito Tephra was first recorded in the central highlands of Papua New Guinea (PNG) in 1971. By the late 1970s, the tephra had been mapped across tens of thousands of square kilometres, traced to its source on Long Island in the Bismarck Sea, and linked to pyroclastic density currents in the Matapun Beds on the island. With a tephra fall volume >10 km³, this eruption was clearly one of the 10 largest globally in the past 600 years. Although almost certainly in the AD 17th or 18th centuries, determining just when this VEI 6 eruption occurred has proved difficult. Whether this eruption occurred before or after William Dampier sailed past, named, described and drew a profile of Long Island has been debated also. Dating the Long Island eruption has implications for assessing the rate of revegetation of the island, recolonisation of the island and caldera lake, Jared Diamond’s theory of ‘supertramp’ birds, correlations between major eruptions and ice core chronologies, the constant rate of supply model of ²¹⁰Pb concentration in lake sediments, reservoir effects in Lake Kutubu in the southern highlands of PNG, the timing of a prehistoric phase of agriculture in the Western Highlands and the longevity of oral histories recording a taim tudak (time of darkness) when sand fell from the sky, houses collapsed, crops were ruined and people died. This paper reviews early dates based on radiocarbon, ²¹⁰Pb and paleosecular magnetic variation, historical reports and genealogical dates from oral histories, and speculation based on tree ring and ice core evidence. Analyses of 10 new radiocarbon dates from the Matapun Beds on Long Island are reported. Our best estimate places the eruption between 1651 and 1671 AD with a 95.4% probability and between 1655 and 1665 AD with a 68.2% probability.

Keywords

eruption implications Long Island oral history Papua New Guinea radiocarbon dating volcanic forcing

Introduction

Across the highlands of Papua New Guinea (PNG), local people from a variety of language groups preserved oral histories of a taim tudak (time of darkness) when sand fell from the sky, crops were ruined, a few houses collapsed and people died. Tephrostratigraphic investigations at numerous sites led to the naming of the taim tudak deposit as ‘Tibito Tephra’ (Blong, 1982).

By the late 1970s, Tibito Tephra had been mapped across an area of tens of thousands of square kilometres of the PNG highlands, and had been traced back to its source on Long Island (5.33°S, 147.1°E), an island in the Bismarck Sea to the north of the PNG mainland (Figure 1). At the same time, it became clear that Tibito Tephra was coeval with major pyroclastic flow deposits on the island first briefly described by G. A. M. Taylor (1953), labelled the ‘pyroclastic mantle’ by Ball and Johnson (1976) and formally named the Matapun Beds by Pain et al. (1981). With the tephra fall volume alone exceeding 10 km³, it was clear that the eruption responsible for this tephra could be classified VEI 6.

Dating the VEI 6 eruption that deposited Tibito Tephra across the highlands and the associated pyroclastic Matapun Beds on Long Island proved to be difficult, though all estimates lay between early 17th and late 19th century. Radiocarbon dates in both areas demonstrated that the eruption occurred at a time when the relationship between radiocarbon years and calendar years was at its most complex and offering a range of interpretations. The evidence provided by the journals of the few European navigators that had traversed the area since the 16th century proved similarly difficult to interpret, and estimates based on genealogies recorded in oral histories suggested more recent dates.

By the 1980s, evidence that this was a VEI 6 eruption encouraged dendrochronologists and ice core scientists to attempt to match cold years and/or sulphate spikes to the eruption, usually without reference to or an understanding of other information that might constrain estimates of the age. As tree ring counting and ice core analysis improved, eruptions elsewhere were associated with particular tree ring and ice core features, and new radiocarbon calibration curves became available, speculation about the year in which the last major eruption of Long Island occurred became a moveable feast – generally believed to be within the latter half of the 17th century, and without any new evidence based on the age of Tibito Tephra or the Matapun Beds.

This contribution reviews the earlier evidence for the age of the eruption, outlines the reasons that dating the eruption as accurately as possible is important and provides the results of a new program of radiocarbon dating of the Matapun Beds.

Background

Long Island lies 130 km due east of Madang and about 55 km north of the nearest point on the PNG mainland (Figure 1). The maximum N-S length of the island is about 27 km and the maximum width about 25 km, with a land area of about 330 km² and a large central caldera. Mt Reaumur, the highest point at 1280 m, is located near the northern end of the island. Remarkably little is known about the climate on the island: Ball and Glucksman (1978) estimated the mean annual rainfall at about 2800 mm with the wettest months with common thunderstorms between December and March, and a dry season extending April to November. Thornton (2001) concurs with earlier authors that the vegetation appears to be an arrested sub-climax of fast growing softwood species dispersed by floating seeds or by birds and bats. Above about 300 m elevation, the forest appears more open. Surface streams are inhibited by the deep porous volcanic deposits across much of the island; prolonged dry seasons and droughts are common.

Details of the European discovery and exploration of Long Island (also known as Arup or Ahrup), PNG, are given in Ball (1982b). Although Abel Tasman sailed past the island in late April 1643, he made no comment on it, possibly believing it was a peninsula of the mainland (Sharp, 1968: 228). Long Island was named by the English navigator William Dampier on 31 March 1700,¹ who also drew a profile of the island. Blong et al. (2016) have shown, after determining the likely location from where Dampier’s profile was drawn, that his view of the island is remarkably similar to the present-day profile (Figure 2), suggesting that the last major eruption of Long Island more likely occurred before AD 1700.

The central freshwater lake (Figure 3) that occupies most of the island was evidently not known to Europeans until 1928 (Ball, 1982b). The lake has a surface area of a little over 90 km², with a surface elevation of about 196 m a.s.l.² Ball and Glucksman (1978) used an echosounder to produce a bathymetric map of the lake; a similar survey by Rabaul Volcanological Survey staff in 1976 resulted in a reasonably similar map (Chris McKee, personal communication, 2014). Our lake bathymetry, based on a DEM most likely derived from the Ball and Glucksman (1978) map, shows that more than 60% of the lake floor (i.e. more than 55 km²) lies below modern sea level with the deepest point at −173 m to the east of the inner island, Motmot (Figure 4). Eruptions have been recorded seven or eight times from Motmot since 1933 and once (1993) from a linear vent system beneath the east central part of the lake (Global Volcanism Program, 2013b).

Based on the bathymetry in Figure 4, the volume of the lake is estimated at about 19.7 km³. The highest point on the caldera rim is approximately 630 m a.s.l. at the northern end; this is almost certainly a ridge of Mt Reaumur. Most of the 40 km long caldera rim lies between elevations of 375 and 500 m a.s.l. If we assume the average height of the rim is 400 m, the caldera volume can be estimated at 40.7 km³; at an average elevation of 500 m, the volume becomes 51.5 km³. Thus, between 30% and 48% of the caldera volume is water-filled.

Immediately after the most recent major eruption, the surface area of the caldera was probably smaller as parts of the caldera walls appear to be eroding. However, as little or no sediment can escape from the lake, the caldera was probably deeper immediately post-eruption. Additionally, Motmot Island and other intra-caldera features have been produced by a number of post-caldera eruptions.

Only the broad details of the volcanic history of Long Island are known. Mt Reaumur (1280 m) at the northern end (the right hand end in Figure 2) and Cerisy Peak (1112 m) at the southern end are ancient basaltic stratovolcanoes, named by the navigator Dumont D’Urville in 1827 (Ball, 1982b; Ball and Johnson, 1976). Prominent satellite cones lie on the southwest flank of Reaumur and east and northwest of Cerisy Peak (Figure 2). Most of the rest of the island is mantled with thick pyroclastic deposits that slope away from the central caldera.

Three major eruptions from the caldera have been identified (Blong et al., 1982; Pain et al., 1981). The earliest of these eruptions is represented by the Kiau Beds, which comprise a thick sequence of airfall ash and lapilli beds overlain by a firm ignimbrite. The outer portion of a completely carbonised log horizontal in the basal beds gave a radiocarbon age of 16,040 ± 270 yr BP (SUA-623). The Kiau Beds have been correlated on geochemical grounds with Ep Tephra widespread in the highlands of mainland PNG (Blong et al., 2017). Two radiocarbon dates from near Kuk (Mt Hagen) provide an age range for Ep tephra between 12,000 and 15,000 radiocarbon years.³

The Kiau Beds are overlain by the Biliau Beds, which form a similar sequence of airfall deposits followed by pyroclastic density currents (PDCs), ranging 1 to 8 m in thickness. A carbonised log in the basal unit provided a radiocarbon age of 3990 ± 110 yr BP (SUA-624) (Pain et al., 1981). The Biliau Beds are believed to be correlated with Kim Tephra widespread in highland PNG and dated at between 3500 and 3800 radiocarbon years.⁴

The uppermost thick eruption sequence on Long Island, the Matapun Beds (Pain et al. 1981), are better exposed and better known. The lowermost unit lies directly on a palaeosol at the top of the Biliau Beds and consists of as much as 2 m of well-bedded coarse airfall ash and pumiceous lapilli. This is overlain by unwelded, poorly bedded PDCs composed of scoriaceous lapilli and non-vesicular blocks in a fine sandy matrix. Accretionary lapilli are common. The PDCs are overlain by up to 1.5 m of discontinuous airfall-bedded ash and lapilli, including accretionary lapilli. On mainland PNG and in the highlands, this eruption deposited Tibito Tephra, mapped across an area of about 94,000 km² (Figure 1), with a volume of more than 10 km³ and geochemically linked to a Long Island source (Blong, 1982).

The relative ‘sizes’ of these three major eruptions are not known with any precision but Ep Tephra appears to be the thickest of the three tephras on the mainland, with Kim Tephra possibly the thinnest and/or least widespread. All three eruptions have been assigned magnitudes M ≥ 6.3 in the LaMeve (Large Magnitude Explosive Volcanic Eruptions) database (http://www.bgs.ac.uk/vogripa/view/controller.cfc?method=lameve). It is likely that all three eruptions contributed to the caldera volume of 40–50 km³ or more, though it is also probable that minor eruptions built new cones and refilled portions of the caldera at various times between the major events. It is not possible to relate caldera size to the volume of each eruption.

While Long Island has produced three major eruptions of global significance in the last 15,000 years, it is the timing of the eruption that produced the Matapun Beds and Tibito Tephra that is the focus of the current contribution.

Significance of the most recent major eruption

Dating of the most recent major eruption of Long Island has implications for several areas of research:

The airfall tephra volume of more than 10 km³ identifies the latest major eruption of Long Island as a VEI 6 eruption. As only 10 or so VEI 6 or greater eruptions have been recorded globally in the last 600 years, the eruption of Long Island is clearly of global significance. In fact, only 56 out of the 7742 in the Smithsonian catalogue (Siebert et al., 2010) of Holocene eruptions have been of this magnitude or greater (roughly 0.7% of all eruptions in the Holocene). Our recent estimates of airfall tephra volume, based on updated and improved methods of volume estimation, suggest the Long Island eruption was larger than the 1883 Krakatau eruption.⁵ A more accurate date will lead to improved estimates of the return periods of such large magnitude events.

Surveys of bird populations on islands in the eastern Bismarck Archipelago show that Long Island, Crown Island and Tolokiwa (Figure 1) have spectacularly high population densities of birds of ‘supertramp’ species, compared with other islands in the region (Diamond, 1974). Supertramps are birds of high overwater dispersal ability but probably low competitive ability. Supertramps are confined to islands placing a premium on dispersal ability – that is, small islands subject to frequent population extinctions, remote islands and islands with recent large volcanic eruptions resulting in defaunation. In fact, Long Island has populations of more supertramp bird species than any other known island in the world (Jared Diamond, 2015, personal communication).

Long Island’s biotic communities appear less developed than expected when compared with islands in the Krakatau group after the 1883 eruption, with the vegetation in an arrested stage of development. Thornton (2001) attributes this at Long Island to the highly porous volcanic soils, poor water retention, low rainfall and frequent extreme seasonal aridity producing slow growth of trees and arrested vegetational development which in turn inhibits maturation of the island’s animal communities.The biota in Lake Wisdom are less developed than that in Lake Dakataua in West New Britain, which formed following a caldera-forming eruption in the AD 7th century (McKee et al., 2011). Ball and Glucksman (1980) hypothesise this results from the shorter time since the creation of the lake and the greater distance of Long Island from sources of colonists (Thornton, 2001).

Tibito Tephra is readily identifiable in numerous mainland PNG lakes and alpine bogs where its use as a tephrostratigraphic marker has contributed to discussions about rates of sedimentation and infilling, nutrient accession and human modification of catchments (Hope et al., 1988). The presence and estimated age of Tibito Tephra in Lakes Egari and Ipea in the highlands has also been crucial to the prominence of the constant rate of supply (c.r.s.) model versus the constant initial concentration (c.i.c.) model of ²¹⁰Pb concentrations in lake sediments (Oldfield et al., 1978). Tibito Tephra has also been used to correlate between lake cores measuring recent palaeomagnetic secular variation (Thompson and Oldfield, 1978).

Recent radiocarbon dating of lake sediments in Lake Kutubu in the southern highlands of mainland PNG (see Figure 1) indicates substantial reservoir effects. Geochemical identification of Tibito and Olgaboli Tephras in the stratigraphic sequence suggests that the reservoir effect amounts to more than 2000 years at the level of Tibito Tephra (Schneider et al., in preparation).

At the Kuk UNESCO World Heritage Site (Golson, 2017) in the Western Highlands (where Tibito Tephra was first identified in 1971), Tibito and other tephra were vital tephrostratigraphic markers elucidating the various phases of agricultural activity at the site. Tibito provides an important marker at the end of Phase 5 of the agricultural history (Golson et al., 2017). The abandonment of wetland cultivation at Kuk took place just before the fall of Tibito Tephra, whereas the most dramatic period of forest disturbance at numerous highland sites occurs immediately after Tibito Tephra fell with evidence of increased erosion (Corlett, 1984). Dating Tibito Tephra also provides crucial evidence for the development of agroforestry based on Casuarina oligodon, increased forest disturbance, increased soil erosion and the appearance of raised-bed garden forms and ditching features in the archaeological record (Haberle, 1998a).

During the 1970s, it became apparent that widespread stories of a taim tudak (time of darkness) across the highlands of PNG were indigenous recognition of the fall of tephra; in most locations across at least 25 language groups, Tibito Tephra was the youngest identifiable tephra (Figure 1). Blong (1982) showed that many of the legends were essentially accurate accounts of the probable consequences of a 10- to 100-mm-thick fall of volcanic ash, except in those aspects relating to the likely duration and timing of the ash fall. Although many accounts included a genealogical basis for the timing of the ash fall (Blong, 1982; Lacey, 1990; Mai, 1981), and the ash fall and legend had earlier been attributed to the 1883 eruption of Krakatau (Watson, 1962), more thorough examination revealed that the oral histories of the tephra fall had survived more or less intact for far longer than the few generations proposed by many of the story tellers. Accurate dating of the fall of Tibito Tephra improves our understanding of the longevity of these oral histories, perhaps with implications about the timing of events portrayed in oral histories elsewhere.

Across much of the highlands, the only historical time that can be accurately determined is the first arrival of European man (usually in or after the 1930s). However, some legends refer to events before or after the taim tudak; for example, several accounts state that the people concerned lived elsewhere when the ash fall occurred or that they migrated to or from another place before or after the ash fall. Several of the oral histories collected about the taim tudak refer to sweet potato vines or tubers (e.g. Nelson, 1971), implying the presence of the food crop at the time of the eruption.⁶

Along the Madang coast, oral histories refer to a major eruption and disappearance of a high island west of Long Island. These legends consistently place the Yomba eruption (Mennis, 2007; Global Volcanism Program, 2013a) at least two generations before the Long Island eruption (Mennis, 2007). While some authorities believe that Yomba was located at or near Hankow Reef (location on Figure 1) northwest of Long Island, an eruption in Holocene times from this source seems very unlikely (Eli Silver, 2014, personal communication; Silver et al., 2009). Although Blong (1982) implied that Yomba may have been the source volcano for Olgaboli Tephra, widespread across the highlands of PNG, this tephra, dated at 980–1180 yr cal BP, probably erupted from Karkar Island (Schneider et al., 2017).

In each of the above examples, improved dating of the most recent major eruption of Long Island provides crucial constraints on the timing, on the rate of operation and/or rate of change of environmental processes, and on our understanding of the human environment in PNG.

Age estimates for the most recent major eruption of Long Island

The timing of the most recent major eruption of Long Island has been of interest for more than a hundred years. Figure 5 summarises views about the age of the eruption based on previous publications (see also Appendix 1). Most of the estimates lie in the range early–mid-17th century to 1883, the latter because the ash fall that produced the taim tudak story in part of the Eastern Highlands was assumed to result from the cataclysmic eruption of Krakatau in 1883 (Watson, 1962).

Age estimates made before the mid-1970s are based entirely on oral histories (O in Figure 5) and nearly all suggest AD post-1700. Bamler (1912), Nurton (1932a) and Carey’s (1938) accounts are based on stories about the eruption of Long Island itself. Vicedom and Tischner (1943, Volume 3, p. 50), early Catholic missionaries among the recently ‘discovered’ highland populations around Mt Hagen, recorded the taim tudak story in the 1930s, the first record of the very widespread accounts of the fall of Tibito Tephra from highland PNG.⁷ These recorders heard a local story about a ‘time of darkness’ when sand fell from the sky, crops, animals and many people died and houses collapsed. These oral accounts frequently included genealogies which referred to an ancestor who was alive at the time of the event; an age estimate can then be made by assigning 25–30 years to each generation before that of the story teller (Blong, 1979b; 1982). The dashed line on Figure 1 encloses the area in which we are certain Tibito Tephra is the youngest identifiable volcanic ash.

Geological (G in Figure 5) or volcanological evidence of a ‘recent’ major eruption on Long Island was first recognised by G. A. M. Taylor in 1953, although the two samples he later submitted for radiocarbon dating (NZ R332 and R517) produced ‘Modern’⁸ results and it has not been possible to identify accurately the locations near the Lake Wisdom crater wall from where these samples were obtained. Three samples collected by Ian Hughes (ANU-1125-1127) from the coast southwest of Poin Kiau (Figure 3) in the early 1970s were also not stratigraphically secure, but the stratigraphy along this coast exposes only Matapun Beds (Pain et al., 1981).

Blong (1982) used four radiocarbon age estimates from Long Island and six from the Kuk prehistoric site in the Western Highlands of PNG. Kuk samples were corrected to allow for (constant) sedimentation rates in the Kuk swamp, the means of the various samples were pooled and the nascent correction curves were examined, to decide that the Stuiver (1978) curves were likely to be the most reliable.⁹ It was concluded: These calibrated radiocarbon dates place the time of darkness, the fall of Tibito Tephra and the eruption of Long Island fairly firmly in the early–mid 17th century. However, consideration of 2σ each side of the mean ages indicates a possibility that the eruption occurred in the early 18th century or, for that matter, in the late 16th century (Blong 1982: 193). Polach (1981) reached essentially the same conclusion using a slightly different route based on the same initial samples.

Blong also examined genealogical estimates of the timing of the occurrence of the time of darkness based on Mai’s 24 interviews (Mai, 1981) in Enga Province and a further 29 accounts collected from across the highlands (Blong, 1981), concluding that most oral histories suggested that the eruption occurred in the mid-18th century or more recently, even when the ‘gap’ between generations was assumed to be an unlikely long 30 years.

Oldfield and co-workers reported an estimated age for Tibito Tephra of ~AD 1860 based on ²¹⁰Pb concentration versus depth curves in Lakes Egari and Pipiak (see locations on Figure 1), based on a constant initial concentration model. Using the more realistic constant rate of supply model provided an age range of AD 1800–1840 (with a mean of AD 1814) (Oldfield et al., 1978). Subsequently, improved measurement techniques for ²¹⁰Pb and ²²⁶Ra using direct gamma assay altered this estimate to AD 1680–1690 (Oldfield et al., 1980), by curve fitting through points fixed at 130–160 years by ²¹⁰Pb and ¹⁴C dates >1000 years (²¹⁰Pb has a half-life of 22.6 years and this updated estimate is beyond the range of the ²¹⁰Pb technique). Thompson and Oldfield (1978) also compared palaeomagnetic secular variation in the Lake Ipea sediments with AD post-1650 secular variation based on all observatory records, concluding that Tibito Tephra was deposited AD pre-1700.

Other evidence of the possible age of the eruption included Dampier’s profile of Long Island drawn in AD 1700 (Figure 2) and his description that: ‘The Trees appeared very green and flourishing …’ (Williamson, 1939). The two lines of evidence suggest that the eruption must have occurred at least 20–30 years prior to his visit (i.e. before AD 1670–1680 to explain the unchanged shape of the island and to allow time for revegetation). Limited historical evidence provided by subsequent European navigators suggested that the eruption could not have taken place after about AD 1800 (Ball, 1982b; Blong, 1982: 193).

After weighing all the available evidence and favouring the radiocarbon dates, Blong (1982: 194) concluded that in the mid-17th century, AD 1630–1670 was the most likely time for the last major eruption of Long Island and the associated events to have occurred. This age estimate was recorded in the Smithsonian Institution’s Volcanoes of the World (1981) as 1660 ± 20 years based on uncorrected radiocarbon years and remains recorded that way today (Global Volcanism Program, 2013b).

Interpreting the Long Island radiocarbon results, and those dating the fall of Tibito Tephra across the highlands of PNG, has not been straightforward as illustrated in Figures 6 and 7. For example, ANU-1126 is a sample of the outer rings of a tree trunk carbonised in a PDC of the Matapun Beds near Poin Kiau (one of the samples collected by Ian Hughes). When the sample was analysed in 1973, the result, 230 ± 75 yr BP (Before Present = 1950), implied a calendar year age of AD 1720, with a one standard deviation range of 1645–1795 and a two standard deviation range of 1570–1870. Figure 6 illustrates that there is not a simple one-to-one correspondence between radiocarbon years and calendar years and that one of the periods most difficult to interpret is AD 1660–1950.

During the last few decades, each new version of the calibration curves gave slightly different results, with none of them resolving satisfactorily one of the central issues – whether the last major eruption occurred before or after William Dampier’s description of Long Island in AD 1700. As Figure 2 shows, there is excellent agreement between Dampier’s 1700 profile of Long Island and a modern profile, but this similarity was less convincing until recently when it became possible to pinpoint the location from where Dampier’s profile had been drawn (Blong et al., 2016). As almost all the slopes on Long Island facing east and between the peaks of Reaumur and Cerisy are mantled with thick deposits of Matapun Beds (Pain et al., 1981), it is now reasonably certain that Dampier’s profile postdates the last major eruption; that is, the eruption occurred at a time before AD 1700.

Dampier also provided a description of the vegetation on Long Island in 1700:

The 31st in the Forenoon we shot in between 2 Islands, lying about 4 Leagues asunder; with Intention to pass between them. The Southernmost is a long Island, with a high Hill at each End; this I named Long Island. The Northernmost is a round high Island towering up with several Heads or Tops, something resembling a Crown; This I named Crown-Isle, from its Form. Both these Islands appear’d very pleasant, having Spots of green Savannahs mixt among the Wood-land: The Trees appeared very green and flourishing, and some of them looked white and full of Blossoms. (Dampier, 1729, in Williamson, 1939)

Although there has been subsequent discussion about the state of the vegetation (Ball, 1982a; Ball and Johnson, 1976), it is clear by analogy with the 1883 Krakatau eruption and vegetation recovery that at least two or three decades must have passed between the eruption and Dampier’s voyage (see, for example, Thornton, 2000, 2001).

By the early 1980s, it was evident that sulphate spikes in the Greenland and Antarctic ice caps recorded a sequence of major volcanic eruptions (e.g. Hammer et al., 1980). As ice cores record more or less annual snow deposition and it is reasonably straightforward to identify the signal of large tropical eruptions such as Krakatau (in 1883) and Tambora (in 1815), the search began for large eruptions that might match other sulphate spikes. As large eruptions also produce global cooling, it was possible to match (particularly European) tree ring patterns to eruptions. The early efforts of Zielinski et al. (1994) and Briffa et al. (1998) were certainly cognisant of the arguments limiting the date of Long Island’s most recent major eruption presented above. However, some subsequent ‘volcanic forcing’ publications (VF in Figure 5) failed to recognise that at least 20 or 30 years were necessary after the eruption for revegetation of the island before Dampier’s AD 1700 description; the authors were seemingly unaware of any of the discussions about the age or arguments for and against timing of the eruption. In the light of Dampier’s description of the vegetation, an AD 1675 date for the eruption seems marginally possible and an AD 1695 date quite unlikely.

These issues were compounded by the problem that ice core records were not always annual as seasonal accumulation layers could not always be easily identified, and erosion sometimes removed other layers so that matches between cores were less than perfect, requiring revision. Similarly, the match between tree rings and ice cores was not always error-free (Baillie, 2010). The efforts of Gao et al. (2008) in comparing numerous ice cores from both the Arctic and Antarctic appears to have reduced errors inherent in reconstructions based on small numbers of cores. However, as Figure 6 and Appendix 1 show, the Long Island eruption has been matched to a number of sulphate spikes and to calendar years. While the VEI 6 eruption of Long Island was certainly large enough, sulphurous enough (Schonwalder et al., 2017) and close enough to the equator to produce a recognisable sulphate spike in ice accumulations in both Northern and Southern Hemispheres, it is not yet possible to match the eruption with any particular spike.¹⁰

New ¹⁴C dates

In June 2014, fresh samples of charcoalised logs from the Matapun Beds on Long Island were collected. Several logs were found in multiple locations around the island. All logs were small and did not contain visible rings preventing any wiggle matching attempts. Charcoal analysis determined the logs were from multiple taxa, Cocos nucifera, Pandanus cf. tectorius, Terminalia cf. catappa and Calophyllum inophyllum (M. Prebble, personal communication, 2017). Samples were returned to the Australian National University Radiocarbon Laboratory for subsequent analysis. A total of 10 samples were selected for radiocarbon dating. Subsamples weighing ~5mg were dried, then pretreated using standard acid:base:acid techniques, 1M HCl for 1 hr, multiple 1M NaOH rinses until clear, 1M HCl, then repeated rinses with high purity water (milli-Q). Approximately 1.5 mg of material was converted to CO₂ at 900°C in a sealed quartz tube containing ~60mg CuO and silver. All charcoal samples were >50% carbon indicating well-preserved samples. The resulting CO₂ was purified and converted to graphite in the presence of hydrogen with powdered Fe as a catalyst. The graphite was then measured on the ANU single stage accelerator mass spectrometer (Fallon et al., 2010). All data were corrected using online AMS d¹³C, normalised to Oxalic Acid I and background subtracted using ¹⁴C free coal treated in the same ABA manner. Data are presented according to the recommendations of Stuiver and Polach (1977).

OxCal 4.2 (Bronk Ramsey, 2009a, 2009b) was used to calibrate the radiocarbon ages to calendar ages. Consistent with previous studies (McKee et al., 2015), we used the northern hemisphere IntCal dataset for calibration as PNG lies close to the northern hemisphere and the SHCal dataset is primarily made up from wood south of 35°S. Figure 6 shows that the eruption date falls in a very difficult period for radiocarbon calibration, with multiple calendar ages possible for a conventional radiocarbon age. Performing a standard calibration on the available Long Island material results in a calendar age eruption estimate of ~1400–present (Table 1, Figure 7).

Table 1.

All radiocarbon samples available for Long Island. Includes OxCal 4.2 calibrated calendar ages and outlier modelled ages.

LAB #	Sample ID	Latitude	Longitude	F¹⁴C	±	¹⁴C age	±	from	to	%	from	to	%
NZ 517	NZ 517	−5.334	147.169	0.9754	0.0081	200	65	1644		68.1	1522		95.4
NZ R 332	NZ R 332	−5.282	147.175	0.9864	0.0049	110	40	1690	1925	68.3	1677	1940	95.4
ANU-754A	Blong Sample 11	−5.787	144.332	0.9766	0.0080	190	65	1648		68.3	1527		95.4
ANU-1052	W109	−5.787	144.337	0.9645	0.0074	290	60	1497	1661	68.2	1450		95.3
ANU-1053	W115	−5.787	144.337	0.9706	0.0074	240	60	1523		68.1	1481		95.4
ANU-1054	W148	−5.783	144.337	0.9645	0.0074	290	60	1497	1661	68.2	1450		95.3
ANU-1055	¹⁴C Sample 10	−5.783	144.337	0.9420	0.0074	480	60	1330	1471	68.2	1305	1625	95.4
ANU-1125	L1/2	−5.235	147.05	0.9538	0.0086	380	70	1447	1630	68.2	1425	1648	95.4
ANU-1126	L1/3	−5.229	147.058	0.9718	0.0092	230	75	1522		68.2	1486		95.3
ANU-1127	L1/5A	−5.215	147.076	0.9754	0.0080	200	65	1644		68.1	1522		95.4
ANU-1307	ANU-1307	−5.258	147.169	0.9574	0.0086	350	70	1465	1634	68.2	1432	1664	95.4
ANU-1933	76/W2/C1	−5.779	144.334	0.9633	0.0086	300	70	1490	1655	68.2	1444		95.4
SANU-33412	Matapun 1 (1976)	−5.366	147.038	0.9611	0.0039	320	20	1521	1636	68.2	1492	1643	95.4
SANU-38524	Matapunlog1-1out	−5.362	147.042	0.9771	0.0029	185	25	1666		68.2	1656		95.4
SANU-39421	MatabeachL 1-mid	−5.366	147.032	0.9791	0.0025	260	30	1530	1795	68.2	1520		95.3
SANU-38406	MatabeachLog2-1	−5.366	147.032	0.9674	0.0031	265	20	1650		68.3	1641		95.4
SANU-39420	matapunlog6	−5.362	147.042	0.9738	0.0034	215	30	1534	1663	68.2	1526	1796	95.4
SANU-39423	MatabeachL 1-ins	−5.366	147.032	0.9740	0.0038	210	35	1649		68.1	1644		95.4
SANU-39424	saokoriver_1	−5.252	147.036	0.9672	0.0032	270	30	1525	1664	68.2	1514	1799	95.4
SANU-39425	saokoriver_2	−5.252	147.036	0.9717	0.0031	230	30	1645	1950	68.2	1530		95.4
SANU-39426	saokoriver_3	−5.252	147.036	0.9720	0.0037	230	35	1643		68.3	1526		95.4
SANU-39427	Matapunlog_4	−5.360	147.044	0.9704	0.0032	240	30	1643	1798	68.2	1526		95.5
SANU-39429	PK7	−5.225	147.063	0.9701	0.0031	245	30	1641	1798	68.2	1523		95.5

In order to refine the estimate of the eruption age, Bayesian modelling was performed in OxCal (Bronk Ramsey, 2009b). We used a standard phase model with outlier analysis to determine a date for the eruption at Long Island (Bronk Ramsey, 2009a). The general outlier model was used with a 0.05 probability of any sample being an outlier (Bronk Ramsey, 2009b). We only used the recently collected samples (those with SANU sample numbers) in the phase model as the providence is robust and we are confident that the outer edge of the logs corresponds to a single event. The modelled calibrated ages for each sample are shown in Figure 8. The modelled date for the eruption has a 1 sigma calendar date range of 1649–1665 AD (68.2%) and a 2 sigma range of 1640–1675 AD (92.5%) and 1785–1793 AD (2.9%).

If we add a priori knowledge to the OxCal phase model using the evidence from Dampier’s observations in 1700 and allowing a minimum 20 years for post-eruption vegetation growth (i.e. the eruption had to occur before 1680 AD), we can further constrain the eruption date to 1651–1671 (95.4% probability, Figure 9).¹¹

Conclusion

For nearly the last five decades, the date of the most recent major eruption of Long Island has been of interest, contributing to debates of historical, archaeological, biological, volcanological and climatic interest. In the 1970s, estimates of the timing of the eruption ranged from the mid-17th century to the very early 20th century with all of the post-1800 AD estimates relying on genealogies reported in oral histories (Figure 5). Estimates based on ²¹⁰Pb concentrations and palaeomagnetic secular variations in lake sediments, supported by radiocarbon dates, then moved estimates into the late 17th or early 18th century. William Dampier’s AD 1700 profile of Long Island (Figure 2) and description of the vegetation added to the confusion about the timing of the eruption.

Resolution of the age of the eruption was complicated by the remarkably non-linear relationship between radiocarbon ages and calibrated dates (Figure 6) and the minor changes engendered by each new calibration. Nonetheless, in the 1980s, the Smithsonian Institution recorded the age of the eruption as 1660 ± 20 years (described as ‘uncorrected’ though that was not the case). Undoubtedly, this age range for a VEI 6 eruption stirred the interest of those researching tree rings and ice cores and the suggested matching to a range of dates between 1640 and 1695, usually without knowledge of the constraints imposed by Dampier’s observations.

The 10 new radiocarbon dates reported here further constrain the possible timing, with our best estimate producing a 95.4% probability that the eruption occurred between AD 1651 and 1671.

Further refinement of this estimate will depend on wiggle matching of charcoalised logs exhibiting annual rings in the Matapun Beds or a convincing match to ice core and/or tree ring records.

Footnotes

Appendix 1

Published estimates of the timing of the last major eruption of Long Island and/or the fall of Tibito Tephra. Estimates are presented in order of publication or, where known, the year in which the information was collected.^a Calendar year estimates are often less precise than they appear – often central estimates ±20 years. Not infrequently, there are gaps of 5 or more years between record and publication. Calendar year estimates are then grouped by decades in Figure 5.

Acknowledgements

We gratefully acknowledge the following individuals: Jack Golson who initiated the search for tephra in archaeological sites in the Western Highlands in 1970 and his continuing interest in our efforts to refine the dating; Ian Hughes who first proposed, in the mid-1970s, that Long Island might be the source of Tibito Tephra; Colin Pain who accompanied CMcK and RB to Long Island in 1976 and extended far and wide our understanding of the distribution of Tibito Tephra; Geoff Hope who over more than 40 years cored far-flung lakes and bogs that contained Tibito across the highlands; Mike Prebble for his identification of the tree species in our ¹⁴C sample collection in 2014; and Paul Grasso for his manipulation of spreadsheets. We are extremely grateful also to the staff of the Madang Resort Hotel, Andrew Barter and the crew of the Kalibobo Spirit for making the visit by RB, SF and CMcK to Madang and Long Island in 2014 so pleasant and rewarding. We also want to thank Frank Oldfield and two anonymous reviewers for their valuable comments on the manuscript.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

Notes

References

Baillie

MGL

(2010) Volcanoes, ice-cores and tree-rings: One story or two? Antiquity 84: 202–215.

Ball

(1982a) Annotated Bibliography of References relating to Long island, Papua New Guinea. Records of the Australian Museum 34(9): 527–547.

Ball

(1982b) Long Island, Papua New Guinea – European Exploration and Recorded Contacts to the end of the Pacific War. Records of the Australian Museum 34(9): 447–461.

Ball

Glucksman

(1978) Limnological studies of Lake Wisdom, a large New Guinea caldera lake with a simple fauna. Freshwater Biology 8: 455–468.

Ball

Glucksman

(1980) A limnological survey of Lake Dakataua, a large caldera lake on West New Britain, Papua New Guinea, with comparisons to Lake Wisdom, a younger nearby caldera lake. Freshwater Biology 10: 73–84.

Ball

Johnson

(1976) Volcanic history of Long Island, Papua New Guinea. In: Johnson

(ed) Volcanism in Australasia: A Collection of Papers in Honour of the Late G A M Taylor. Amsterdam: Elsevier, pp. 133–147.

Ballard

Brown

Bourke

et al . (eds) (2005) The Sweet Potato in Oceania: A Reappraisal (Ethnology Monographs 19/Oceania Monographs 56). Sydney, NSW, Australia: The University of Sydney.

Bamler

(1912) Das Erdbeben vom 14 auf den 15 September 1906. Naturhistoriche Gesellschaft Nuremburg Abhandlungen 19: 17–18.

Blong

(1975) The Krakatoa myth and the New Guinea Highlands. Journal of the Polynesian Society 84(2): 213–217.

10.

Blong

(1979a) Huli legends and volcanic eruptions, Papua New Guinea. Search 10: 93–94.

11.

Blong

(1979b) Time of darkness legends from Papua New Guinea. Oral History 7(10): 1–135.

12.

Blong

(1981) Time of darkness: legends and volcanic eruptions in Papua New Guinea. In: Denoon

Lacey

(eds) Oral Tradition in Melanesia. Waigani: University of Papua New Guinea, pp. 141–150.

13.

Blong

(1982) The time of darkness: Local legends and volcanic reality in Papua New Guinea. Canberra, ACT, Australia. ANU Press.

14.

Blong

Kemp

Chen

(2016) Dampier’s view of Long Island, Papua New Guinea. Terrae Incognitae: The Journal for the History of Discoveries 48(2): 139–159.

15.

Blong

Pain

McKee

(1982) Long Island, Papua New Guinea – Aspects of landforms and tephrostratigraphy. Records of the Australian Museum 34(9): 419–426.

16.

Blong

Wagner

Golson

(2017) Tephras at Kuk. In Golson

Denham

Hughes

et al . (eds) Ten Thousand Years of Cultivation at Kuk Swamp in the Highlands of Papua New Guinea (Terra Australis 46). Canberra, ACT, Australia: ANU Press, pp. 117–131.

17.

Bradbury

(2006) Reconstructing the East Asian monsoon response to major volcanic eruptions: A test of model skill with instrumental and paleoclimate data. PhD Thesis, University of Massachusetts Amherst, Amherst, MA.

18.

Briffa

Jones

Schweingruber

et al . (1998) Influence of volcanic eruptions on Northern Hemisphere summer temperature over the past 600 years. Nature 393: 450–455.

19.

Bronk Ramsey

(2009a) Bayesian analysis of radiocarbon dates. Radiocarbon 51(1): 337–360.

20.

Bronk Ramsey

(2009b) Dealing with outliers and offsets in radiocarbon dating. Radiocarbon 51(3): 1023–1045.

21.

Brookfield

(1961) The Highlands People of New Guinea: A study of distribution and localization. Geographical Journal 127: 436–448.

22.

Buckland

Dugmore

Edwards

(1997) Bronze Age myths? Volcanic activity and human response in the Mediterranean and North Atlantic regions. Antiquity 71: 581–593.

23.

Carey

(1938) The morphology of New Guinea. The Australian Geographer 3(5): 3–31.

24.

Casey

(2010) Correlation of solar activity minimum and large magnitude geophysical events. Research Report 1-2010 (preliminary). Orlando, FL: Space and Science Research Center.

25.

Corlett

(1984) Human impact on the subalpine vegetation of Mt Wilhelm, Papua New Guinea. Journal of Ecology 72: 841–854.

26.

Crowley

Quinn

Taylor

et al . (1997) Evidence for a volcanic cooling signal in a 335-year coral record from New Caledonia. Paleoceanography 12(5): 633–639.

27.

Davies

(2012) The geology of New Guinea – The cordilleran margin of the Australian continent. Episodes 35(1): 87–102.

28.

D’Arrigo

Wilson

Tudhope

(2009) The impact of volcanic forcing on tropical temperatures during the past four centuries. Nature Geoscience 2: 51–56.

29.

Diamond

(1974) Colonization of exploded volcanic islands by birds: The supertramp strategy. Science 184: 803–806.

30.

Du Toit

(1969) Misconstruction and problems in communication. American Anthropologist 71: 46–53.

31.

Fallon

Fifield

Chappell

(2010) The next chapter in radiocarbon dating at the Australian National University: Status report on the single stage AMS. Nuclear Instruments and Methods in Physics Research Section B 268: 898–901.

32.

Gao

Robock

Ammann

(2008) Volcanic forcing of climate over the past 1500 years: An improved ice core based index for climate models. Journal of Geophysical Research. DOI: 10.1029/2008JD010239.s

33.

Garrett-Jones

(1979) Evidence for changes in Holocene vegetation and lake sedimentation in the Markham Valley, Papua New Guinea. PhD Thesis, Australian National University, Canberra ACT, Australia.

34.

Glasse

(1963) Bingi at Tari. Journal of the Polynesian Society 72: 270–271.

35.

Global Volcanism Program (2013a) Hankow Reef (251041). In: Venzke

(ed.) Volcanoes of the World. v.4.5.6. Washington, DC: Smithsonian Institution. Available at: http://volcano.si.edu/volcano.cfm?vn=251041 (accessed 13 October 2017).

36.

Global Volcanism Program (2013b) Long Island (251050). In Venzke

(ed.) Volcanoes of the World. v.4.5.6. Washington, DC: Smithsonian Institution. Available at: http://volcano.si.edu/volcano.cfm?vn=251050 (accessed 13 October 2017).

37.

Golson

(1997) From horticulture to agriculture in the New Guinea highlands: A case study of people and their environments. In Kirch

Hunt

(eds) Historical ecology in the Pacific Islands. New Haven, CT: Yale University Press, pp. 39–50.

38.

Golson

(2017) An Introduction to the investigations at Kuk Swamp. In Golson

Denham

Hughes

Swadling

Muke

(eds) Ten thousand years of cultivation at Kuk Swamp in the highlands of Papua New Guinea (Terra Australis 46). Canberra, ACT, Australia: ANU Press.

39.

Golson

Denham

Hughes

et al . (eds) (2017) Ten thousand years of cultivation at Kuk Swamp in the highlands of Papua New Guinea. Terra Australis 46). Canberra, ACT, Australia: ANU Press.

40.

Haberle

(1998a) Dating the evidence for agricultural change in the highlands of New Guinea: The last 2000 years. Australian Archaeology 47: 1–19.

41.

Haberle

(1998b) Late Quaternary vegetation change in the Tari Basin, Papua New Guinea. Palaeogeography, Palaeoclimatology, Palaeoecology 137: 1–24.

42.

Hauk

(2009) Dendrochemistry: Seeing the forest through the trees. PhD Thesis, Pennsylvania State University in Nuclear Engineering, University Park, PA.

43.

Hammer

Clausen

Dansgaard

(1980) Greenland ice sheet evidence of post-glacial volcanism and its climatic impact. Nature 288: 230–235.

44.

Hofstede

van de Wal

RSW

Kaspers

et al . (2004) Firn accumulation records for the past 1000 years on the basis of dielectric profiling of six cores from Dronning Maud Land, Antarctica. Journal of Glaciology 50(169): 279–291.

45.

Hope

Gillieson

Head

(1988) A comparison of sedimentation and environmental change in New Guinea shallow lakes. Journal of Biogeography 15(4): 603–618.

46.

Humphreys

(1998) A review of some important soil studies in Papua New Guinea. Papua New Guinea Journal of Agriculture, Forestry and Fisheries 41(1): 1–19.

47.

Johnson

(2013) Fire mountains of the islands – A history of volcanic eruptions and disaster management in Papua New Guinea and the Solomon Islands. Canberra, ACT, Australia: ANU Press.

48.

Johnson

Taylor

GAM

Davies

(1972) Geology and petrology of Quaternary volcanic islands off the north coast of New Guinea (Australian Bureau of Mineral Resources, Geology and Geophysics Record 1972/21). Available at: https://d28rz98at9flks.cloudfront.net/12702/Rec1972_021.pdf.

49.

Jones

Briffa

Schweingruber

(1995) Tree-ring evidence of the widespread effects of explosive volcanic eruptions. Geophysical Research Letters 22(11): 1333–1336.

50.

Lacey

(1990) Temps perdu et temps vécu: Cross-cultural nuances in the experience of time among the Enga. Pacific Studies 13: 77–102.

51.

Leahy

(1969) Exploration into Highland New Guinea, 1930–1935 (ed Goodale

). Tuscaloosa, AL: University of Alabama Press.

52.

Xie

S-P

Cook

et al . (2013) El Nino modulations over the past seven centuries. Nature Climate Change 3: 822–826.

53.

Luterbacher

Rickli

Xoplaki

et al . (2001) The late Maunder Minimum (1675–1715) – A key period for studying decadal scale climatic change in Europe. Climatic Change 49: 441–462.

54.

Mai

(1981) The ‘time of darkness’ or Yuu Kuia. In Denoon

Lacey

(eds.) Oral tradition in Melanesia. Waigani: University of Papua New Guinea, pp. 125–140.

55.

McKee

Baillie

Reimer

(2015) A revised age of AD 667–669 for the latest major eruption of Rabaul. Bulletin of Volcanology 77. DOI: 10.1007/s00445–015–0954–7

56.

McKee

Neall

Torrence

(2011) A remarkable pulse of large-scale volcanism on New Britain Island, Papua New Guinea. Bull. Volcanology 73: 27–37.

57.

Meggitt

(1973) The Sun and the Shakers: A millenarian cult and its transformations in the New Guinea Highlands. Oceania 44(1): 1–37; (2): 109–126.

58.

Mennis

(2007) Tsunami on the north coast of Papua New Guinea. TsuInfo Alert 9(4): 4–8.

59.

Naval Intelligence Division (1945) Pacific Islands. Volume IV: Western Pacific (New Guinea and Islands Northward) (Geographical Handbook Series B. R. 519C: 42, 6). Washington, DC: Naval Intelligence Division.

60.

Nelson

(1971) Disease, demography and the evolution of social structure in highlands New Guinea. Journal of the Polynesian Society 80: 204–216.

61.

Nurton

(1932a) Long Island and S. East portion of Sub Division 7 – Rai Coast. August–September 1932. Madang District. Australian Archives series no. A7034/28

62.

Nurton

(1932b) Patrol to Long Island for purposes of collecting Head Tax and Census. Madang District. Australian Archives series no. A7034/206.

63.

Oldfield

Appleby

Battarbee

(1978) Alternative ²¹⁰Pb dating: Results from the New Guinea Highlands and Lough Erne. Nature 271: 340–342.

64.

Oldfield

Appleby

Thompson

(1980) Palaeoecological studies of lakes in the highlands of Papua New Guinea: 1. The chronology of sedimentation. Journal of Ecology 68(2): 457–477.

65.

Oldfield

Worsley

Appleby

(1985) Evidence from lake sediments for recent erosion rates in the highlands of Papua New Guinea. In Douglas

Spencer

(eds) Environmental Change and Tropical Geomorphology. London: Allen & Unwin, pp. 185–196.

66.

Osipov

Khodzer

Golobokova

et al . (2013) High resolution 900yr volcanic and climatic record from the Vostokara, East Antarctica. The Cryosphere Discussions 7: 1961–1986.

67.

Pain

Blong

(1979) The distribution of tephras in the Papua New Guinea highlands. Search 10: 228–230.

68.

Pain

Blong

McKee

(1981) Pyroclastic deposits and eruptive sequences on Long Island, Papua New Guinea. In: Johnson

(ed.) Cooke-Ravian Volume of Volcanological Papers, Geological Survey of Papua New Guinea Memoir 10. Port Moresby, PNG: PNG Geological Survey, pp. 101–107.

69.

Palmer

van Ommen

Curran

MAJ

et al . (2001) High-precision dating of volcanic events (A.D. 1301–1995) using ice cores from Law Dome, Antarctica. Journal of Geophysical Research 106(D22): 28089–20095.

70.

Polach

(1981) Pyroclastic deposits and eruptive sequences of Long Island. Part 2: Radiocarbon dating of Long Island and Tibito tephras. In: Johnson

(ed.) Cooke-Ravian Volume of Volcanological Papers, Geological Survey of Papua New Guinea Memoir 10. Port Moresby, PNG: PNG Geological Survey, pp. 108–113.

71.

Roullier

Benoit

McKey

et al . (2013) Historical collections reveal patterns of diffusion of sweet potato in Oceania obscured by modern plant movements and recombination. Proceedings National Academy of Sciences 110(6): 2205–2210.

72.

Schneider

Alloway

Blong

et al . (2017) Stratigraphy, age and correlation of two widespread Late Holocene tephras preserved within Lake Kutubu, Southern Highlands Province, Papua New Guinea. Journal of Quaternary Science 32: 782–794.

73.

Schneider

Haberle

Fallon

et al . (submitted to Quaternary Science Reviews) Evaluating the radiocarbon reservoir effect in Lake Kutubu, Southern Highlands, Papua New Guinea.

74.

Schonwalder

Costa

Blong

(2017) The atmospheric S budget emitted by the ‘Tibito Tephra’ eruption (mid-17th Century), Long Island volcano, Papua New Guinea. Portland, OR: International Conference for Volcanology and Chemistry of the Earth’s Interior Conference Abstract VO43A-144).

75.

Sharp

(1968) The Voyages of Abel Janszoon Tasman. Oxford: Clarendon Press.

76.

Siebert

Simkin

Kimberley

(2010) Volcanoes of the world. 3rd Edition. Washington, DC: Smithsonian Institution, University of California Press.

77.

Sigl

McConell

Toohey

et al . (2014) Insights from Antarctica on volcanic forcing during the Common Era. Nature Climate Change 4: 693–697.

78.

Silver

Day

Ward

et al . (2009) Volcano collapse and tsunami generation in the Bismarck Volcanic Arc, Papua New Guinea. Journal of Volcanology and Geothermal Research 186: 210–222.

79.

Simkin

Siebert

(1994) Volcanoes of the World. 2nd ed. Washington, DC: Smithsonian Institution.

80.

Sniderman

JMK

Finn

Denham

(2009) A late-Holocene palaeoecological record from Ambra crater in the highlands of Papua New Guinea and implications for agricultural history. The Holocene 19(3): 449–458.

81.

Stuiver

(1978) Radiocarbon timescale tested against magnetic and other dating methods. Nature 273: 271–274.

82.

Stuiver

Polach

(1977) Discussion: Reporting of 14C data. Radiocarbon 19:355–363.

83.

Sullivan

(2009) Middle to Lower Ramu subsistence, household and culture study second revision (Nancy Sullivan and Associates Ltd Consulting Report). Port Moresby, Papua New Guinea: Nancy Sullivan and Associates Ltd.

84.

Taylor

GAM

(1953) Notes on Ritter, Sakar, Umboi and Long Island volcanoes (Australian Bureau of Mineral Resources Records 1953/43). Available at: https://d28rz98at9flks.cloudfront.net/9262/Rec1953_043.pdf.

85.

Thompson

Oldfield

(1978) Evidence for recent palaeomagnetic secular variation in lake sediments from the New Guinea Highlands. Physics of the Earth and Planetary Interiors 17: 300–306.

86.

Thornton

IWB

(2000) The ecology of volcanoes: Recovery and reassembly of living communities. In Sigurdsson

Houghton

McNutt

et al . (eds) Encyclopedia of volcanoes. San Diego, CA: Academic Press, pp. 1057–1081.

87.

Thornton

IWB

(2001) Colonization of an island volcano, Long Island, Papua New Guinea, and an emergent island, Motmot, in its caldera lake. I. General Introduction. Journal of Biogeography 28: 1299–1310.

88.

Traufetter

Oerter

Fischer

et al . (2004) Spatio-temporal variability in volcanic sulphate deposition over the past 2 kyr in snow pits and firn cores from Amundsenisen, Antarctica. Journal of Glaciology 50(168): 137–146.

89.

Tung

K-K

Zhou

(2013) Using data to attribute episodes of warming and cooling in instrumental records. Proceedings of the National Academy of Sciences of the United States of America 110(6): 2058–2063.

90.

Venturati

(2016) Interplay of adverse natural factors as possible cause of decline and abandonment of the Sasso-Simone city-fortress during the Renaissance (AD 1566–1673). In: Bulgarian Geological Society, National Conference with international participation ‘Geosciences 2016’, 7–8 December. Sofia: Bulgarian Geological Society, pp. 141–142.

91.

Vicedom

Tischner

(1943) Die Mbowamb. Monographien zur Volkerkunde, 3 vols. Museum fur Volkerkunder und Vorgeschichte. Hamburg: Friederischen, de Gruyter & Co.

92.

Wahl

Diaz

Smerdon

et al . (2014) Late winter temperature response to large tropical volcanic eruptions in temperate western North America: Relationship to ENSO phases. Global and Planetary Change 122: 238–250.

93.

Watson

(1962) Krakatoa’s echo. Journal of the Polynesian Society 72: 152–155.

94.

Williamson

(1939) A Voyage to New Holland. 1939 Edition. London: The Argonaut Press.

95.

Zielinski

(1995) Stratospheric loading and optical depth estimates of explosive volcanism over the last 2100 years derived from the Greenland Ice Sheet Project 2 ice core. Journal of Geophysical Research 100(D10): 20937–20955.

96.

Zielinski

Mayewski

Meeker

et al . (1994) Record of volcanism since 7000 B.C. from the GISP2 Greenland ice core and implications for the volcano-climate system. Science 264: 948–952.

Significance and timing of the mid-17th-century eruption of Long Island,Papua New Guinea

Abstract

Keywords

Introduction

Background

Significance of the most recent major eruption

Age estimates for the most recent major eruption of Long Island

New 14C dates

Conclusion

Footnotes

Appendix 1

Acknowledgements

Funding

Notes

References

New ¹⁴C dates